Home Theater Forum and Systems banner

21 - 40 of 94 Posts

·
Premium Member
Joined
·
1,562 Posts
Sorry for the confusion. Yes, I inadvertently switched my reference to the midrange horn from MF to MR (mid frequency to midrange). I see that above in post-8 I also used 'mid-horn'. I will try to be more consistent and use 'MR' for 'midrange' in the future.

So for the MR horn: Set the HPF to LR-12 and set the LPF to LR-24.

The logic here is that the 2 horns are likely rolling off their low frequencies near the XO frequency at a But-12 or greater rate without any 4080 XO filter applied. We likely only need a LR-12 (or But-12) HPF to increase that rate so that the total acoustical rate approaches LR-24. After we see what we actually get we can adjust the filter as needed. The closer we get to a symmetrical LR-24 XO the greater the opportunity to get close phase tracking of the drivers throughout the XO range. That will result in a stable frontal lobe and also a more consistent SPL off-axis (in the horizontal plane). These are relatively minor considerations to overall SPL at the LP, but logically provide the best opportunity to achieve a relatively wide and stable sound field at the LP and likely result in more stable imaging and spatial effects.
 

·
Registered
Joined
·
99 Posts
Discussion Starter #22
Ok John,
Thanks for clearing that up for me, now I understand it. Ok will make the changes and also do you want the TW inverted? I will try and do this tonight.
 

·
Premium Member
Joined
·
1,562 Posts
Yes, we can invert the TW's now. Their polarity will then adhere to the conventional setting.

In the upcoming timing alignment step we will determine the needed polarity of the MR and W to assure all drivers are working together through the XO ranges.
 

·
Registered
Joined
·
99 Posts
Discussion Starter #25 (Edited)
Ok John,
I hope I got it correct this time, ok the LPF is 24db on the W, the MR LPF is 24db and the HPF is 12db per your instructions. TW is still 12db. The W is xover at 400hz, the MR xover is from 400hz to 6000hz, TW xover is 6000hz. The TW is inverted to neg. polarity...
 

Attachments

·
Premium Member
Joined
·
1,562 Posts
Okay,
First the issues:
> Your mic calibration file does not appear to be very accurate. The overall roughness is too great and the shape also appears to deviate from the expected profile a little too much. I don't think it is bad enough that it will be a major problem though. We normally adjust the house curve a little to taste anyway so it may wash out due to that. I only mention it in case you a highly motivated to achieve high accuracy measurement to work from. A calibration from Cross-Spectrum Labs would provide higher confidence.
> It appears something went wrong in the loopback timing of the HF horn. It was delayed more than the other drivers were. Make sure your loopback cable is connected in channel 1 output to input and that the input level is properly set on that timing channel. I was able to see the evidence of the issues due to the the timing channel crosstalk into the measurement channel. I made the needed adjustment to correct the HF Horn impulse position so the delay analysis below is still good.

Analysis of the XO filter shapes:
> Both acoustic XOs have reasonably good symmetry with these XO filter settings. The lower XO is shaped a little better than the upper one. The roll-off of the HF horn appears to be lower in frequency and shallower than ideal, but it is still very usable. We could confirm the frequency was set correctly and possibly increase the HPF frequency a little and also increase the slope to 18 dB/octave and see it that helps. The phase tracking looks pretty good though so the impact of any changes is not likely to be significant.

Acoustic XOs:
Acoustic XOs.jpg

Timing Analysis:
Since the filter settings look reasonable and I was able to correct the misplaced HF Horn Impulse it was possible to go ahead with step 2 to determine the delay adjustment needed for close phase tracking.
> HF Horn: It's the reference so no changes to its delay. Its polarity is also correct now.
> MR Horn: Invert this driver and increase its delay by 0.86 ms.
> W: Invert this driver and leave the delay unchanged.

Phase tracking Lower XO:
Phase Tracking Lower XO.jpg

Phase tracking Upper XO:
Phase Tracking Upper XO.jpg

EQ Considerations:
With this mic position it is possible to EQ the range above the floor / ceiling bounce frequencies. With the appropriate window settings we can we can safely EQ down to maybe 500 Hz. At this mic distance there is only minor roll-off of high frequencies. In the EQ chart below I did set a very minor slope to the high end reflecting my personal preferences in my setup. You may want to change this as per your preferences. Below 500 Hz it is better to take an average of several measurements around the LP area. It is normally fine to EQ the range below 200 Hz or so to a single LP measurement if you prefer. The mid range from 200-500 is best done with and average measurement and taking care not to be too aggressive. Many prefer to EQ to account for the baffle step effect, but otherwise not use any other EQ in that range. Below I offer a starting point for EQ above 200 Hz as the response needs some EQ to account for the direct sound SPL. The bass EQ must be done from the LP so you can adjust the Woofer level and EQ the response there accordingly. Just with the rough EQ setting I noted below the sound quality is likely to be very good. Both channels should be EQ'ed identically above 500 Hz. Below that you can experiment to see what works best for you. I have had good luck using identical EQ for both channels down to approximately 250 Hz in my setup.

EQ settings:
EQ.JPG

Predicted Response With EQ >500Hz
Predicted SPL with EQ.jpg
 

·
Registered
Joined
·
99 Posts
Discussion Starter #27
Okay,
First the issues:
> Your mic calibration file does not appear to be very accurate. The overall roughness is too great and the shape also appears to deviate from the expected profile a little too much. I don't think it is bad enough that it will be a major problem though. We normally adjust the house curve a little to taste anyway so it may wash out due to that. I only mention it in case you a highly motivated to achieve high accuracy measurement to work from. A calibration from Cross-Spectrum Labs would provide higher confidence.

John that's strange I guess it is a new mic with cal.

> It appears something went wrong in the loopback timing of the HF horn. It was delayed more than the other drivers were. Make sure your loopback cable is connected in channel 1 output to input and that the input level is properly set on that timing channel. I was able to see the evidence of the issues due to the the timing channel crosstalk into the measurement channel. I made the needed adjustment to correct the HF Horn impulse position so the delay analysis below is still good.

Oops, I didn't have the loopback cable connected. Maybe the cause of some bad info?



Analysis of the XO filter shapes:
> Both acoustic XOs have reasonably good symmetry with these XO filter settings. The lower XO is shaped a little better than the upper one. The roll-off of the HF horn appears to be lower in frequency and shallower than ideal, but it is still very usable. We could confirm the frequency was set correctly and possibly increase the HPF frequency a little and also increase the slope to 18 dB/octave and see it that helps. The phase tracking looks pretty good though so the impact of any changes is not likely to be significant.

Acoustic XOs:
View attachment 142370

Timing Analysis:
Since the filter settings look reasonable and I was able to correct the misplaced HF Horn Impulse it was possible to go ahead with step 2 to determine the delay adjustment needed for close phase tracking.
> HF Horn: It's the reference so no changes to its delay. Its polarity is also correct now.
> MR Horn: Invert this driver and increase its delay by 0.86 ms.
> W: Invert this driver and leave the delay unchanged.

I do have some delay already put into the HF and MR horns, is this gonna be a problem?

Phase tracking Lower XO:
View attachment 142378

Phase tracking Upper XO:
View attachment 142386

EQ Considerations:
With this mic position it is possible to EQ the range above the floor / ceiling bounce frequencies. With the appropriate window settings we can we can safely EQ down to maybe 500 Hz. At this mic distance there is only minor roll-off of high frequencies. In the EQ chart below I did set a very minor slope to the high end reflecting my personal preferences in my setup. You may want to change this as per your preferences. Below 500 Hz it is better to take an average of several measurements around the LP area. It is normally fine to EQ the range below 200 Hz or so to a single LP measurement if you prefer. The mid range from 200-500 is best done with and average measurement and taking care not to be too aggressive. Many prefer to EQ to account for the baffle step effect, but otherwise not use any other EQ in that range. Below I offer a starting point for EQ above 200 Hz as the response needs some EQ to account for the direct sound SPL. The bass EQ must be done from the LP so you can adjust the Woofer level and EQ the response there accordingly. Just with the rough EQ setting I noted below the sound quality is likely to be very good. Both channels should be EQ'ed identically above 500 Hz. Below that you can experiment to see what works best for you. I have had good luck using identical EQ for both channels down to approximately 250 Hz in my setup.

EQ settings:
View attachment 142394

Predicted Response With EQ >500Hz
View attachment 142402
I will wait to hear your reply before I change anything.

Jay
 

·
Premium Member
Joined
·
1,562 Posts
Oops, I didn't have the loopback cable connected. Maybe the cause of some bad info?
Yes, that is a problem for step-2. Fortunately, at full zoom with the 'normalize' box cleared, we can see the weak crosstalk of the timing channel in the measurements. REW also detected this crosstalk in the timing channel and thus placed 2 of the 3 measurements accurately in time. REW didn't detect the HF Horn timing properly though. It was possible to manually do that however, so the proper positioning was manually restored for this analysis. As a result, this was 'good data' for Step-2 (delay) analysis as well as the intended Step-1 (XO Filter) analysis. This was just a heads-up reminder that for future measurements requiring timing accuracy that the cable must be in place for reliable results.

I do have some delay already put into the HF and MR horns, is this gonna be a problem?
No, the adjustment is just a change to the current setting so just add 0.86 ms to the current MR Horn delay setting. Leave the delay of the W and HF-horn unchanged. Don't forget to invert the polarity of the W and MR-Horn. I forgot to mention the recommended EQ changes should be applied to the left and right inputs in the 4080, not to the 4080 driver outputs. If you had some EQ already applied you can just add these 4 recommended EQ filters to those current settings.

Next Action:
It is now appropriate to remeasure to confirm the results of the changes. To do that just make the recommended changes to delay, polarity and EQ and then remeasure the 3 drivers with loopback timing engaged. I also recommend measuring the other channel with all the same settings, but with the appropriate mic positioning for that channel. When the predicted changes are confirmed by the new measurements and both channels look similar then we are ready for Step-3.

If you post the new left and right channel driver measurements, I can help confirm the Step-2 results.

Step-3 involves moving the mic to the LP for bass EQ adjustment.

Step-4 would involve adjusting the house curve as needed to your preference.

Beyond the 4 EQ filters recommended here as a starting point to confirm the XO settings and provide basic direct sound control, I will leave additional EQ work for your investigation and implementation.
 

·
Registered
Joined
·
99 Posts
Discussion Starter #29
Ok John,
That is good news, I will make the changes and re measure and post results.
Now you are saying that after I make these changes and measure to move on to the right channel and do as I did for the left?

Jay
 

·
Premium Member
Joined
·
1,562 Posts
Yes, make the settings on the other channel the same as this channel. Then place the mic in the correct place for that channel and take the same set of 3 measurements. We just want to confirm the 2 channels are well matched.
> If there are significant differences between channels then we need to resolve that.
> If there are significant differences between the predicted response and the actual measured response then we need to resolve that.
> If all is as expected then this step is complete.
 

·
Registered
Joined
·
99 Posts
Discussion Starter #31
John,
I only had time to do the Left channel tonight sorry, I think I have it all correct have a look. Also I just left everything setup for tomorrow, if I got this correct I will do the Right channel tomorrow during the day.
 

Attachments

·
Premium Member
Joined
·
1,562 Posts
Thanks for the data.
Well, the measured result does not match the predicted result so there is a problem someplace. The EQ looks good and the drivers polarity is correct per the provided instruction. The delay of the MR horn is not correct however resulting in poor phase tracking and an SPL dip in the XO range. On close examination of the timing crosstalk the 3 drivers reference times do not fall on top of each other as expected. This suggests the REW loopback timing may not be working properly for some reason.

I suggest:
> Let's first confirm loopback timing is setup and working properly. Measure the MR Horn 6 times in succession still using the same 20-20k sweep. Post the file so I can confirm the impulse location of each measurement is identical.
> If the measurements do not confirm repeatability, we need to investigate further to determine where the problem is.
> When the repeatability test is successful we can we restart Step-2 delay settings from the beginning. that is, repeat the 3 driver measurements on the left channel to determine new delay recommendations.
> Since we will be starting Step 2 over you can change the mic height as needed to be on your line of sight from the LP to a point on the baffle between the 2 horns. [You indicated to me the 36" height I originally suggested was too low for your high seating position so you can make that height change now.
 

·
Premium Member
Joined
·
1,562 Posts
I reviewed the data and it looks okay for repeatability. Measurement #4 and #6 deviated very slightly from the other 4. Both may have been due to some ambient room noise during the measurements. I see no major problem with loopback timing repeatability in this data. The MR-horn timing however in the May-31 and Jun-01 measurements is significantly different that lead to this problem. I cannot be sure which timing is correct. Each of the 3 sessions is also slightly different in timing. Possibly that is because windows is locking onto an REW session using a slightly different latency??

The safest path forward is...
> Use one REW session for all the following measurements on the left channel. Nothing should change except the driver selected for measurement.
Measure in the following order: (do not deviate from this order)
> HF-horn
> MR-horn
> W
> HF-horn
> MR-horn
> W
> HF-horn
> MR-horn
> W

That is, measure each of the 3 drivers and then repeat that same series 3 times so there are 9 total measurement in the one file. If the file is then too large to post here you can then break the original file into 2 or 3 smaller files or even email the entire file to me directly.

This way I can confirm again that the timing is repeatable and if necessary I can also avoid using any odd measurement that appears to be corrupted in time or by noise. I will then provide new driver delay adjustments for the 4080 as needed.
 

·
Premium Member
Joined
·
1,562 Posts
Okay the analysis is complete.
The timing of the MR-horn was wrong. It did agree with the Post-31 so it is consistent with the last set of measurement. I can only conclude that my original analysis in Post-26 was erroneous.

The adjustment needed is to the MR-horn. It needs a delay increase of 0.58 ms.

The polarities are all correct as measured.

The charts of Post-26 still apply. Below is the overall phase response.
Overall SPL-Phase.jpg

Now we should confirm this analysis.
Measure Left Channel with the new MR delay:
> HF-horn
> MR-horn
> W

Measure Right Channel with the new setting and the appropriate mic position:
> HF-horn
> MR-horn
> W
 

·
Registered
Joined
·
99 Posts
Discussion Starter #39
Ok John,
That's good news!!! Now adding the delay of 0.58ms plus the 4.083 that is already in would give me a total delay of4.663 ms correct?
 
21 - 40 of 94 Posts
Top